Surgical removal of malignant disease constitutes one of the most common and effective therapeutic for primary treatment for cancer. Resection of all detectable malignant lesions results in no detectable return of the disease in approximately 50% of all cancer patients1 and may extend life expectancy or reduce morbidity for patients in whom recurrence of the cancer is seen. Not surprisingly, surgical methods for achieving more quantitative cytoreduction are now receiving greater scrutiny.
Resection of all detectable malignant lesions results in no detectable return of the disease in approximately 50% of all cancer patients and may extend life expectancy or reduce morbidity for patients in whom recurrence of the cancer is seen. Given the importance of total resection of the malignant lesions, it is beneficial to ensure that the malignant lesions are accurately and completely identified. Identification of malignant tissue during surgery is currently accomplished by three methods. First, many tumor masses and nodules can be visually detected based on abnormal color, texture, and/or morphology. Thus, a tumor mass may exhibit variegated color, appear asymmetric with an irregular border, or protrude from the contours of the healthy organ. A malignant mass may also be recognized tactilely due to differences in plasticity, elasticity or solidity from adjacent healthy tissues. Finally, a few cancer foci can be located intraoperatively using fluorescent dyes that flow passively from the primary tumor into draining lymph nodes. In this latter methodology, fluorescent (sentinel) lymph nodes can be visually identified, resected and examined to determine whether cancer cells have metastasized to these lymph nodes.
Despite the recognition of the importance of removal of tumor and the availability of certain identification techniques for visualizing tumor mass, many malignant nodules still escape detection, leading to disease recurrence and often death. Thus, there is a need for improved tumor identification. This motivation has led to introduction of two new approaches for intraoperative visualization of malignant disease. In the first, a quenched fluorescent dye is injected systemically into the tumor-bearing animal, and release of the quenching moiety by a tumor-specific enzyme, pH change, or change in redox potential is exploited to selectively activate fluorescence within the malignant mass. In the second approach, a fluorescent dye is conjugated to a tumor-specific targeting ligand that causes the attached dye to accumulate in cancers that over-express the ligand's receptor. Examples of tumor targeting ligands used for this latter purpose include folic acid, which exhibits specificity for folate receptor (FR) positive cancers of the ovary, kidney, lung, endometrium, breast, and colon, and DUPA, which can deliver attached fluorescent dyes selectively to cells expressing prostate-specific membrane antigen (PSMA), i.e. prostate cancers and the neovasculature of other solid tumors. Beneficially, one folate-targeted fluorescent dye (folate-fluorescein or EC17) has been recently tested intra-operatively in human ovarian cancer patients. In this study, ˜5× more malignant lesions were removed with the aid of the tumor-targeted fluorescent dye than without it, and all resected fluorescent lesions were confirmed by pathology to be malignant.
Conventional fluorescent techniques use probes in the visible light spectrum (˜400-600 nm), which is not optimal for intra-operative image-guided surgery as it is associated with a relatively high level of nonspecific background light due to collagen in the tissues. Hence the signal to noise ratio from these conventional compounds is low. Moreover, the absorption of visible light by biological chromophores, in particular hemoglobin, limits the penetration depth to a few millimeters. Thus tumors that are buried deeper than a few millimeters in the tissue may remain undetected. Moreover ionization equilibrium of fluorescein (pKa=6.4) leads to pH-dependent absorption and emission over the range of 5 to 9. Therefore, the fluorescence of fluorescein-based dyes is quenched at low pH (below pH 5).
For example, the potential use of EC17 dye for a more widespread use in optical imaging for the characterization and measurement diseased tissue in a clinical setting has been hampered by the major drawback of that the attached dye (fluorescein) emits fluorescence in the visible range. This makes EC17 and related dyes poor for in vivo use in tissues because tissues typically autofluoresce strongly in the visible range, and light penetrates tissue poorly. Moreover, EC17 (folate-ethelenediamine-fluorescein isothiocynate) consists a thiourea linker. It is well known that thiourea compounds have low shelf life due to the instability of the thiourea linkage. Thus, a compound such as EC17 is not optimal for use in optical imaging because of this unstability and the related decomposition of the decomposition of thiourea bridge.
The combination of light absorption by hemoglobin in the visible light spectrum (<600 nm) and water and lipids in the IR range (>900 nm), offers an optical imaging window from approximately 650-900 nm in which the absorption coefficient of tissue is at a minimum. A suitable alternative to dyes that emit light in the visible range would be to develop dyes that can be used in the near infra red (NIR) range because light in the near infrared region induces very little autofluorescence and permeates tissue much more efficiently. Another benefit to near-IR fluorescent technology is that the background from the scattered light from the excitation source is greatly reduced since the scattering intensity is proportional to the inverse fourth power of the wavelength. Low background fluorescence is necessary for highly sensitive detection. Furthermore, the optically transparent window in the near-IR region (650 nm to 900 nm) in biological tissue makes NIR fluorescence a valuable technology for in vivo imaging and subcellular detection applications that require the transmission of light through biological components.
While the use of light in the NIR range for deeper tissue imaging is preferable to light in the visible spectrum, the NIR imaging dyes currently used in the art suffer from a number of challenges and disadvantages such as a susceptibility to photobleach, poor chemical stability, absorbance and emission spectra that fall within the same range as many physiological molecules (resulting in high background signal and autofluorescence). Moreover, most of the NIR dyes are not stable during the synthesis, especially conjugating to a ligand with an amine linker, leading to multiple unwanted side products. Therefore, taking ligand-targeted NIR imaging agent for clinic can be expensive. Thus, current imaging methods that utilize NIR fluorescent probes are not effective in deep tissue imaging (>5 mm from the surface), in quantifying fluorescence signal in mammalian tissues, or in production cost that increase preclinical-to-clinical translational time.
Two promising approaches to fluorescence-guided surgery are currently under intense investigation for use in the clinic. In one method, an activatable NIR fluorescent probe, which is minimally fluorescent in the steady state due to its proximity to an attached quencher, becomes highly fluorescent upon release of the quencher in malignant tissue. One of the most commonly used release mechanisms involves incorporation of a peptide sequence between the dye and the quencher that can be specifically cleaved by a tumor-enriched protease (i.e. cathepsins, caspases and matrix metalloproteinases). A major advantage of this strategy lies in the absence of fluorescence in tissues that lack the activating enzyme, allowing tissues along the excretion pathway (e.g. kidneys, bladder, liver) to remain nonfluorescent unless they fortuitously express the cleaving enzyme. Such tumor-activated NIR dyes can also generate substantial fluorescence in the tumor mass as long as the malignant lesion is enriched in the cleaving protease and the released dye is retained in the tumor. The major disadvantage of this methodology arises from the poor tumor specificities of many of the relevant hydrolases (most of which are also expressed in healthy tissues undergoing natural remodeling or experiencing inflammation). Moreover, the abundance of the desired proteases may vary among tumor masses, leading to slow or no activation of fluorescence in some malignant lesions and rapid development of fluorescence in others.
Thus, there remains a need for a dye substance that can be used to specifically target diseased tissue and has increased stability and brightness for use in vivo for tissue imaging.